Abstract: As technological progress accelerates and China's demographic structure shifts,human capital has taken on an increasingly central role in economic and social development. Over the past decade,a large number of Chinese cities have followed the relaxation of hukou restrictions with sizable talent attraction policies of their own. This paper provides a systematic study of the resulting talent competition among Chinese cities through the lens of talent subsidy policies,documenting its historical development,its temporal and spatial stylized facts,and the strategic competitive interaction among local governments. We first conduct a comprehensive collection of talent policy documents issued by prefecture-level cities nationwide and construct a city-level talent subsidy index covering the period 2008 to 2019. We find that talent attraction policies across Chinese cities exhibit a pronounced pattern of episodic,phased acceleration over time,a clear east-to-west declining gradient in space,and a significant inverted U relationship with the level of local economic development. We further document a significant peer effect in policymaking at the prefecture level:a city's talent subsidy intensity responds positively to the policies of its potential competitor cities. This finding provides empirical support for the promotion tournament theory of strategic interaction among Chinese local governments. The paper closes with a systematic review of the literature on the effects of talent attraction policies and a discussion of directions for future research. The widely cited starting point of this talent competition is Wuhan's February 2017 program to “retain one million university graduates in five years,” which set off the so-called “talent war” among cities. By December 2019,virtually every prefecture-level city in China had issued some form of talent policy,with more than 85% offering specific talent attraction measures. As eligibility thresholds were gradually lowered from advanced degree holders to ordinary undergraduates,the talent war increasingly took on the character of a broader labor competition,especially against the backdrop of falling fertility. Existing research has largely treated these city-level policies as exogenous shocks to study their effects on local outcomes such as innovation and industrial upgrading. The literature has paid much less attention to the policies themselves:their content,scope,and intensity. Existing measurement strategies rely mainly on binary indicators of policy adoption,counts of policy documents,or subjective text-coding scores,none of which adequately capture the substantive variation in incentive intensity across cities and over time. This paper addresses that gap. The first contribution is empirical. The paper constructs what is,to our knowledge,the most comprehensive panel dataset of city-level talent attraction policies in China,covering 297 prefecture-level cities from 2008 to 2019 and drawing on more than three thousand policy documents collected through systematic multi-channel search. The constructed talent subsidy index incorporates subsidy type,the educational threshold of eligible recipients,and the conditional restrictions attached to disbursement,providing a substantially richer measure than has previously been available. Using this dataset,the paper documents three sets of stylized facts. First,the temporal pattern is one of episodic acceleration rather than gradual diffusion. Fewer than twenty cities per year introduced subsidy policies before 2015. The number of new adopters roughly doubled to forty in 2017 and peaked at fifty-four in 2018. Coverage simultaneously expanded down the education ladder,from doctoral holders toward ordinary undergraduates. Second,the spatial pattern shows a clear east-to-west intensity gradient,with eastern provinces such as Jiangsu and Zhejiang having adopted subsidy policies well before the 2017 wave. By the end of 2019,doctoral-level subsidies covered 252 cities,equivalent to 85% of the sample,while undergraduate-level coverage stood at only 51% and remained concentrated in the east. Third,subsidy intensity exhibits an inverted U relationship with city-level GDP per capita:cities at intermediate levels of economic development offer the most generous subsidies,while both the poorest and the most developed cities are less aggressive. The second contribution is to identify the mechanism behind the rapid escalation of the talent war. The paper attributes it to strategic interaction among local governments operating within China's promotion tournament system. Under fiscal decentralization and politically managed cadre promotion,prefecture officials face strong incentives to compete on outcomes that are rewarded in cadre evaluation,and the 2016 Guideline explicitly incorporated talent work into that evaluation framework. The implication is that high-profile policy moves by a focal city,such as Wuhan in 2017,exert horizontal pressure on neighboring and peer-tier cities to respond in kind,both to protect their local human capital stock from being drawn away and to avoid falling behind in the promotion tournament. The paper tests this mechanism empirically and finds a significant peer effect:a city's subsidy intensity responds positively to the policies of same-tier cities in neighboring provinces. Heterogeneity analysis shows the effect is stronger in ordinary prefecture cities,in eastern regions,in the post-2016 period,and in cities led by younger party secretaries.